Can Avoidant Attachment Change?

12 min read · Reviewed against current psychological research

You’ve probably decided this is just who you are. Some people are built for closeness and you aren’t, and there’s a certain relief in that conclusion, because it means you can stop failing at something you were never going to be good at. Maybe a partner has told you that you’ll never change. Maybe you’ve said it about yourself, in a tone that sounds like acceptance but is closer to a locked door.

The evidence says otherwise, and it’s more interesting than a simple yes. Avoidant attachment can absolutely change, and there’s real data on how, how fast, and, crucially, on the things people think will change it that reliably don’t. That last part matters most, so it’s worth reading to the end even if you already believe the answer is yes.

The short answer, with the data

Two encouraging findings up front.

Avoidance naturally declines with age. Longitudinal research tracking people from adolescence into their seventies finds that avoidant attachment gradually decreases across the lifespan. The drift is already, quietly, in your favour. (Anxious attachment behaves differently, peaking in young adulthood and then falling.)

Therapy moves the needle. Meta-analyses of attachment-focused therapy report moderate to large effect sizes for improvements in attachment security, along with gains in emotional regulation and relationship satisfaction. Longitudinal data show those changes persist beyond the end of treatment, provided the person stays genuinely engaged with it.

So the honest headline is: not fixed, not fated, and demonstrably movable.

The part that gets left out

Here’s the calibration almost nobody includes, and it’s the reason so many people conclude change is impossible after trying.

Attachment security shifts fairly often, roughly half of major life events produce an immediate change. But most people drift back to their baseline within about six to forty months, and only around a quarter of those events produce a lasting move toward security.

Read that carefully, because it’s genuinely useful. It means change happens easily and reverts easily. A new relationship, a good year, a period of feeling unusually open, all of these can shift you, and then the old pattern quietly reasserts itself. Which tells you exactly what the variable is: spontaneous change usually doesn’t stick. Deliberate change does. The difference between a temporary softening and an actual restructuring is whether you did anything on purpose.

What does not work

This is the most valuable section here, because these four things are what most people try, and none of them will do it.

Willpower. Deciding to be more open is not a mechanism. Deactivation isn’t a choice you’re making; it’s a protective reflex that fires below conscious awareness, and you cannot out-decide a reflex.

Insight alone. Knowing you’re avoidant, understanding exactly why, being able to name your deactivating strategies with impressive fluency, none of that changes anything by itself. As one clinician puts it: an X-ray doesn’t set the bone. You can stare at it all day.

Finding the right partner. A better partner improves your experience. It does not restructure your attachment system. That leads to the big one.

A patient partner waiting it out. This one needs saying plainly, because a lot of people are quietly banking on it. A secure, patient partner genuinely helps, they create room to move toward connection. But patience does not change the underlying structure. It changes the experience of one specific relationship while the pattern stays intact for the next one. Attachment travels with the person, not the relationship.

The obstacle that’s specific to avoidance

Now the honest difficulty, and it explains why avoidant people so often don’t seek help.

Anxious attachment hurts loudly. The anxious person is in visible, felt distress, which is unpleasant but has one enormous advantage: it supplies motivation. They know something is wrong because they can feel it constantly.

Avoidance is different, because the strategy is working. The whole function of deactivation is to prevent you from feeling attachment distress, and it succeeds. You genuinely do feel fine. So the obvious question, why would I fix something that doesn’t hurt?, is not stupid at all. It’s exactly what a well-functioning suppression system produces.

The catch is that the research shows the cost is still being paid. People high in avoidance frequently show elevated physiological stress during attachment-related distress while sincerely reporting that they feel nothing. The bill is real. It just isn’t itemized where you can see it. And the suppression tends to fail eventually, under stress, exhaustion, or after a relationship ends, which is why so many avoidant people feel almost nothing during a breakup and get hit hard months later.

What actually works

Attachment-focused therapy. This is the most reliable route, and the approach should match the pattern: dismissive-avoidance and fearful-avoidance need different starting points, which is why it’s worth knowing which one you are. Modalities with support here include EFT, EMDR, and IFS. Fearful-avoidance in particular often needs trauma-informed care.

Repeated corrective experience. Security isn’t installed by understanding; it’s accumulated through evidence. Over months and years, repeatedly experiencing that closeness didn’t cost you what you expected slowly rewrites the old model: maybe I can rely on someone. Maybe having a need isn’t dangerous.

A coherent story. One of the strongest predictors of earned security isn’t a good childhood, it’s the ability to describe your history coherently, with balance rather than either dismissal or being overwhelmed by it. Note that dismissing the past (“it was fine, I don’t think about it”) is itself an avoidant move, and it’s the specific thing that gets in the way.

Time. Meaningful shifts often show up within six to twelve months of consistent therapy plus real-world practice. Deeper restructuring, earned security, generally takes years. Anyone promising faster is selling something.

If you’re the partner reading this

A lot of people arrive at this question hoping the answer is yes so they can keep waiting. So, honestly:

Yes, they can change. And no, you cannot do it for them. Your love is not a treatment. Your patience is not a mechanism. You can make this relationship gentler, and that’s worth something real, but you cannot restructure another adult’s nervous system by being understanding enough for long enough.

There’s a further trap worth knowing about. When both partners are insecurely attached, which is extremely common, the relationship often becomes the place where both people’s worst fears get confirmed rather than corrected. Their withdrawal proves your fear of abandonment; your pursuit proves their fear of being too much for someone. That’s the anxious-avoidant trap, and it’s the opposite of a corrective experience.

Change requires the avoidant person to want it and to do something about it. “They might change someday” is not a plan, and it isn’t a reason to keep going indefinitely while your own needs go unmet. Whether to stay while someone isn’t yet working on themselves is a legitimate decision, but make it as a decision, not as a bet.

The takeaway

Avoidant attachment is not a life sentence. It softens with age on its own, and it responds well to deliberate work. What it does not respond to is willpower, insight, the right partner, or someone else’s patience, and if those are the things you’ve been counting on, the failure wasn’t yours, you were using tools that don’t do this job.

The genuine difficulty is that avoidance doesn’t hurt enough to force the issue. The system is doing precisely what it was built to do, which is keep you from feeling the thing you’d need to feel in order to want it different. So the first move isn’t opening up. It’s simply being willing to consider that the calm might be costing you something you can’t see, and that the person you’d be without it isn’t a weaker version of you, but a less guarded one.

Where to go from here depends on what you want to change. Avoidant attachment is the fuller description of the pattern, and the complete guide to attachment styles shows what secure looks like from where you are standing. If the change you want is behavioral rather than internal, how you communicate under pressure shifts faster than the underlying wiring does, and telling a boundary from a wall is often where the honest work starts. Earned security is one of the clearer things therapy helps build, so choosing an approach deliberately is worth the time.

Frequently asked questions

Sources

1. Longitudinal research on attachment change across the lifespan (samples spanning adolescence to older adulthood): avoidant attachment gradually decreases from adolescence onward; anxious attachment peaks in young adulthood and then declines. See also Chopik W, et al., Attachment security and how to get it, Social and Personality Psychology Compass.
2. Evidence on life events and attachment change: approximately half of major life events produce immediate shifts in attachment security, with most people returning to baseline within roughly 6–40 months and about a quarter of events producing enduring movement toward security.
3. Meta-analytic evidence that cognitive-behavioural and emotion-focused therapies improve attachment security (moderate to large effect sizes), with gains in emotional regulation and relationship satisfaction persisting with continued engagement.
4. Clinical literature on what does and does not change avoidant attachment: willpower, insight alone, partner selection, and a secure partner’s patience do not restructure attachment; “attachment style travels with the person, not the relationship.” Attachment-focused therapy (EFT, EMDR, IFS), extended corrective relational experience, and time do.
5. Main M, and the Adult Attachment Interview literature on earned security: coherence of narrative, rather than the content of early memories, predicts secure functioning. See also Roisman GI, Earned-Secure Attachment Status in Retrospect and Prospect.
6. Mikulincer M, Shaver PR. Deactivating strategies; elevated physiological stress in avoidant individuals alongside low self-reported distress; suppression failing under high cognitive load.